Current Server Time: 23:05 (Central Europe)

#apertus IRC Channel Logs

2019/06/20

Timezone: UTC


01:15
rektide_
joined the channel
01:19
rektide_
was looking at recorder/monitors for my camera & was thinking how much i wanted open source hardware to be able to do SDI, so i showed up here to just say hi & bow before how much awesome work ya'll have done
01:21
Bertl
you're very welcome!
01:39
rektide_
changed nick to: rektide
02:28
BAndiT1983
changed nick to: BAndiT1983|away
04:53
Bertl
off to bed now ... have a good one everyone!
04:53
Bertl
changed nick to: Bertl_zZ
05:19
illwieckz_
joined the channel
05:22
illwieckz
left the channel
07:31
niemand
joined the channel
07:31
niemand
left the channel
07:31
niemand
joined the channel
07:32
sebix
joined the channel
07:49
niemand
left the channel
08:02
sebix
left the channel
08:18
BAndiT1983|away
changed nick to: BAndiT1983
08:58
illwieckz_
left the channel
09:06
se6astian|away
changed nick to: se6astian
10:48
BAndiT1983
changed nick to: BAndiT1983|away
10:56
Umori_
left the channel
10:59
Umori
joined the channel
11:34
Umori
left the channel
11:38
Umori
joined the channel
12:19
se6astian
changed nick to: se6astian|away
13:54
Bertl_zZ
changed nick to: Bertl
13:54
Bertl
morning folks!
13:57
BAndiT1983|away
changed nick to: BAndiT1983
14:36
sebix
joined the channel
14:57
se6astian|away
changed nick to: se6astian
15:01
BAndiT1983
changed nick to: BAndiT1983|away
15:17
RexOrCine|away
changed nick to: RexOrCine
15:36
sebix
left the channel
15:48
se6astian
changed nick to: se6astian|away
15:55
se6astian|away
changed nick to: se6astian
15:55
se6astian
left the channel
15:55
se6astian|away
joined the channel
15:56
se6astian|away
changed nick to: se6astian
15:56
se6astian
left the channel
15:56
RexOrCine
left the channel
15:56
Nira|away
left the channel
15:56
BAndiT1983|away
left the channel
15:56
philippej
left the channel
16:00
Nira|away
joined the channel
16:00
BAndiT1983|away
joined the channel
16:00
BAndiT1983|away
changed nick to: BAndiT1983
16:00
philippej|away
joined the channel
16:00
philippej|away
changed nick to: philippej
16:00
RexOrCine|away
joined the channel
16:00
se6astian
joined the channel
16:00
RexOrCine|away
changed nick to: RexOrCine
16:59
Nira|away
changed nick to: Nira
17:03
danieel
left the channel
17:09
danieel
joined the channel
17:14
se6astian
changed nick to: se6astian|away
17:14
se6astian|away
changed nick to: se6astian
17:25
Y_G
joined the channel
17:45
Y_G
Hi BAndiT1983
17:48
Y_G
I have a response value which is exceeding 1024 ,so should I increase the response buffer or do have any other solution
18:22
se6astian
Hi Y_G, do you have a minute?
18:29
Y_G
hi se6astian ,yes tell me
18:46
se6astian
did BAndiT1983 already chat with you about the wiki documentation of the control daemon CLI examples?
18:46
se6astian
https://wiki.apertus.org/index.php/Control_Daemon#CLI
18:52
Y_G
Yes ,I have went through these ,He asked me to extend these with the functionalities added during the task
18:53
Y_G
I will do once BAndiT1983 approves of the code
18:58
se6astian
What I wanted to propose is adding more examples of commands that are already possible now and do not rely on extending the code
18:58
se6astian
but at the same time making sure they work as expected by testing on the remote beta
18:58
se6astian
in particular chaning image sensor registers
18:58
se6astian
*changing
18:59
se6astian
is that something you could dig into?
19:05
Bertl
off for now ... bbl
19:05
Bertl
changed nick to: Bertl_oO
19:08
Y_G
Checked the code ,there are six functions that are allowed with the un-extended code : `get` and `set` analog_gain,digital_gain ,config_register
19:21
se6astian
yes
19:21
se6astian
the image sensor has 128 register blocks
19:22
se6astian
so examples how to read and set some example registers would be very handy
19:22
se6astian
did you take a look at the image sensor datasheet already?
19:27
Y_G
Yes ,I have gone through it
19:29
se6astian
great
19:30
se6astian
then picking out a few key image sensor parameters and providing examples how to get set them with the control daemon CLI would be great
19:30
se6astian
with full examples of the commands but also of the way the return values are provided
19:30
se6astian
and some explanation what the return values mean exactly and how to interpret them
19:31
Y_G
So do you want specific functions for it for ex "DaemonCLI image_sensor set adc_range 10",or just document what register to set for adc_range
19:31
se6astian
is that something you could work on?
19:31
Y_G
yes definitely,it's a part of my proposal .Will work on it
19:32
se6astian
great
19:32
se6astian
"DaemonCLI image_sensor set adc_range 10" will not work
19:33
se6astian
as you need to provide register index and value
19:33
se6astian
so "just document what register to set for adc_range" is mandatory
19:35
Y_G
if we have a function for adc_range it will automatically know which register to set.
19:36
Y_G
for ex. "DaemonCLI image_sensor set analog_gain 2" ,This automatically sets the register for analog_gain
19:38
Nira
changed nick to: Nira|away
19:47
se6astian
you are right
19:48
se6astian
my focus was on config_register
19:48
se6astian
so "just document what register to set for adc_range" is what I meant
19:48
se6astian
there are two ways how to do the same thing
19:49
se6astian
either its an abstraction already in place
19:49
se6astian
like changing analog gain
19:49
se6astian
or you can look up which image sensor register corresponds to the analog gain
19:49
se6astian
and set the image sensor register directly
19:49
se6astian
thats what I want to have documented on the wiki
19:49
se6astian
and how to do it
19:55
aombk
left the channel
19:55
Y_G
Ok will have a look on what registers look important and will update them accordingly.
19:56
Y_G
Will require for some time for this ,as I have to meet timeline for the first evaluation.
20:35
se6astian
sure
20:35
se6astian
any questions let me know
20:46
RexOrCine
Test
20:48
se6astian
leaving the axiom office now
20:49
se6astian
changed nick to: se6astian|away
21:00
RexOrCine
changed nick to: RexOrCine|away
21:04
BAndiT1983
hi Y_G
21:05
BAndiT1983
what is the problem with 1024 bytes? how is it exceeding?
21:05
BAndiT1983
second thing about index and value for registers, everything is there
21:05
BAndiT1983
the data packet format was extended some while ago and can transfer 2 parameters, setters should use 1 or both, depending on the case
21:07
BAndiT1983
also setters and getters for registers are there also for quite a while, but i had no time to check it, as we had connection problems to HDMI, afterwards my job took some time and will still take a lot of time in the next weeks
21:10
aombk
joined the channel
21:23
Y_G
There is a functionality that requires to capture information about all pac1720 ,The output is as follows : "https://pastebin.com/FDjZNZw8"
21:24
Y_G
The response value is greater than 1024 bytes
21:24
BAndiT1983
pastebin is not longer there
21:24
BAndiT1983
if it exceeds this size, then the approach cannot be correct
21:26
Y_G
https://pastebin.com/MNSUb2EL
21:26
BAndiT1983
ehm, are you really trying to send this data inside flatbuffers packet?
21:28
BAndiT1983
this is exactly the opposite what the daemon communication should do
21:28
BAndiT1983
it would be the task of daemon CLI to format data for output on screen, but the daemon communication should be as small as possible
21:31
Y_G
I have individual functions for getting each row too . But to get the overview of all the sensors would help the user
21:31
Y_G
Even if formating is left to daemon CLI ,it's still a lot of data here
21:32
BAndiT1983
i don't see a necessity to do the formatting on the daemon side
21:32
BAndiT1983
goal is to keep daemon from crashing
21:33
BAndiT1983
have you wrote down the processes and goals, so i can analyse the steps you want to take?
21:35
Y_G
Had a discussion with Bertl this Monday and Tuesday about the goals of breaking down pac1720info.
21:36
BAndiT1983
and what was decided?
21:38
Y_G
My takeaway from that was it would help to have functions for voltages of individual sensors , and the power consumption per rail
21:38
BAndiT1983
ok, but then i don'T understand why you want to create info output in the daemon
21:39
BAndiT1983
there could be a periodical job to grab values, when some script or CLI asks for them, then the cached value is delivered
21:40
BAndiT1983
so oyu would have like a method for 5V_NORTH, 5V_NORTH_AMP or so
21:40
BAndiT1983
*you
21:40
Y_G
It was majorly a test to check I2CAdapter ,to see if it was working properly.Which I suppose does work properly now
21:41
BAndiT1983
if you think it would help to send multiple things, then you have to evaluate how to send proper array and how to use it without hassle while developing modules and while using CLI
21:43
BAndiT1983
another thing, please avoid placing long methods in if/else, you can grab the variables which are different and set values, afterwards call the method using it
21:45
BAndiT1983
smbus.h should be moved to 3rdParty folder, as it seems it was copied from somewhere else
21:45
BAndiT1983
also smbus.cpp of course
21:45
BAndiT1983
what about newer kernels, do we really require this SMBus methods?
21:46
BAndiT1983
word of caution, looks like you've renamed smbus.c to smbus.cpp, first thing is naming, according to C++ it should be SMBus.cpp
21:47
BAndiT1983
second thing is probable problems, because you have to use special keywords sometime to mix C++ and C
21:47
BAndiT1983
-probable +possible
21:49
Dev_
joined the channel
21:51
Dev_
Hello BAndiT1983 ,
21:51
Y_G
I can try without SMBus methods ,should not be a problem.
21:51
BAndiT1983
hi Dev_
21:52
BAndiT1983
Y_G, it'S about evaluating what is the state nowadays, as google was not clear at some points
21:53
Y_G
Have tried going through articles ,all of them suggested to use smbus commands instead of simple read/write
21:53
Y_G
For newer kernels could only find one workaround ,that said just include the functions that are missing from smbus.c
21:54
BAndiT1983
was the reason named, why they were deprecated?
21:57
Y_G
Couldn't find anything on that
21:57
BAndiT1983
https://www.kernel.org/doc/html/latest/driver-api/i2c.html
21:57
BAndiT1983
it only states that I2C supports more things than SMBus, but about the methods it'S not clear
21:59
Dev_
BAndiT1983: I have changed the sharing settings of document . Can u able to check it now
22:00
BAndiT1983
Dev_, yes
22:00
BAndiT1983
will check your allocator tomorrow or while weekend, but have seen quite some problems
22:01
BAndiT1983
have commented on trello and also in the other GDoc file
22:01
Dev_
yes , I saw your comment about IAllocator on trello ,
22:01
Dev_
i will change that
22:02
illwieckz
joined the channel
22:02
BAndiT1983
have you understood why i have commented it?
22:06
BAndiT1983
so, have to leave for today, work awaits tomorrow
22:06
BAndiT1983
if any questions are there, then write here or on trello, will try to do proper code review tomorrow
22:06
BAndiT1983
changed nick to: BAndiT1983|away
22:13
Dev_
The IAllocator is abstract class which contains methods that can be extended for different usecases,
22:16
Dev_
we would be having different implementation for Allocate(), Deallocate() methods for different cases and we can extend the IAllocators methods easily
22:17
Dev_
i will add IAllocator again
22:17
Dev_
working on your comments on Gdoc document now ..
22:48
Dev_
left the channel
23:00
comradekingu
joined the channel
23:12
illwieckz_
joined the channel
23:15
illwieckz
left the channel
23:46
Y_G
left the channel